On Tue, Nov 4, 2008 at 7:52 PM, Alan Coopersmith <Alan.Coopersmith at sun.com> wrote: > Alan Coopersmith wrote: >> - Is there an increase or decrease in application performance >> when using the xcb-ified libX11? > > Looks like the latest git commits fix a big performance drop: > >> Furthermore, the handoff approach provides great improvements to >> performance. Results from x11perf's XNoOp test, which represented the worst >> case for the lock-based Xlib/XCB: >> >> Traditional Xlib: average 19100000/sec >> Lock-based Xlib/XCB: average 3350000/sec >> Handoff-based Xlib/XCB: average 17400000/sec > > They also remove the locking assertions that caused compatibility > problems for things such as Java versions with incorrectly built > private copies of libXinerama. > > (Further details in the commit notes on > http://cgit.freedesktop.org/xorg/lib/libX11/commit/?id=54e5c0941b0ded1628d559a9f0a3451ea96c299b > ) > > Seems like we'd want the upcoming libxcb & libX11 releases when > we're actually ready to integrate. > > -- > -Alan Coopersmith- alan.coopersmith at sun.com > Sun Microsystems, Inc. - X Window System Engineering
No, thanks, WHY should I (or another external person) ever work on the fox-gate again? At the end everything is ready and nobody moves it upstream. I hardly even get a response regarding this. A year back, when the fox-gate was set up, it had been without any question that SPARC-Xorg would be used for one or more of "Sun's distros". Now I get lessons in how to think, what to expect and how to read/interpret emails. Assuming I could afford it to start working on xcb (another month for free), who promises me that the same doesn't happen again? It is ridiculous to think, I would have taken the time to integrate the SPARC stuff into the fox-gate just for fun in its own sense, without expecting Sun to use that stuff in Indiana, SXCE and Solaris 10. And I don't believe anybody, that he or she in fact thought this. Could I go back in time, I would have laughed at Sun when they initially set up opensolaris.org . But over time not even anyhing of the initial good spirit has survived, not a small bit. Today all that opensolaris is, is a commercial "product". For stuff like that I would not have wasted months and months and months of my time, tens of thousands of $$$ for direct bills for the required testing hw and electricity and well beyond 100K of money that I didn't earn, because instead of doing anything for my career I had been a dumb "OpenSolaris enthusiast believer". http://blogs.barrons.com/techtraderdaily/2008/10/31/sun-no-guidance-no-restructuring-no-turnaround/?mod=yahoobarrons Also read the comments section (bullet proof glass for JS, 700K) The X11 group doesn't even get the money to buy a coffee machine, yet millions and hundreds of millions are wasted elsewhere. Actually billions if you consider overpriced aqisitions like Cobalt (Sept. 2000) and StorageTek in 2005. I had been very patient with my sympathy for that company, but everything has an end. This top-management is ill-judged. Or call it "incompatible with reality". Or worse. Martin Bochnig
